Bilkent University
Department of Computer Engineering


Discovery of Structural Variations using Linked-Read Technology


Fatih Karaoğlanoğlu
MS Student
Computer Engineering Department
Bilkent University

Alterations in DNA that affect more than 50 base pairs are referred to as structural variations. High throughput sequencing technologies are being employed for the detection of such events since their inception. There are many tools that can discover SNP’s and short indels, however, accurate characterization of large segmental duplications and balanced rearrangements remains to be an unsolved problem. Contemporary algorithms fail to accurately predict large inversion breakpoints and there is no algorithm to predict the insertion locus of large interspersed segmental duplications. Recently, Linked-Read sequencing technologies such as 10X Genomics were introduced. This technology attaches barcodes that carry long range information to the short reads. We propose a method to characterize direct and inverted interspersed segmental duplications and inversions that are >100 Kbp using Linked-Read sequencing data. We use split-molecule signature derived from linked reads along with read pair and molecule depth signatures to make precise predictions.


DATE: 09 April, 2018, Monday, CS590 & CS690 presentations begin at @ 15:40